Remote Sensing for Agricultural Greenhouse Gas Flux Models: Advanced Multispectral Sensor Requirements

Guy Serbin, InuTeq LLC, guy.serbin@gmail.com (Presenter)
Earle Raymond Hunt, USDA ARS, raymond.hunt@ars.usda.gov
Craig Daughtry, USDA-ARS, craig.daughtry@ars.usda.gov
Martha Anderson, USDA-ARS, manderson@hydrolab.arsusda.gov

Agricultural greenhouse gas (GHG) fluxes are of major concern, because different agricultural practices are either major GHG sources or sinks. Remote sensing offers the ability to rapidly map agricultural GHG flux model parameters, and thus, allow for effective GHG monitoring and management. While hyperspectral sensors that cover the visible (0.4 – 0.7 μm), near-infrared (NIR, 0.7 – 1.2 μm), and shortwave infrared (SWIR, 1.2 – 2.5 μm), in conjunction with thermal infrared (TIR, 10.3 – 12.5 μm) portions of the electromagnetic spectrum allow for these measurements, the data storage and processing requirements can be prohibitive at national and international scales, particularly with frequent revisit times and at medium spatial resolution. An advanced multispectral system containing 14 narrow spectral bands is ideal for agricultural monitoring. Important parameters include live vegetation cover, chlorophyll content, vegetation water content, non-photosynthetic vegetation cover (NPV, e.g., crop residues and dry grasses), evapotranspiration (ET), and vegetation stress, plus bands for atmospheric correction. While current and planned missions can acquire some of these parameters, none are suited for remote sensing of NPV, which is crucial in mapping tillage practices, rangeland health, and brush fire hazards, and modeling soil organic carbon dynamics and soil erosion and water quality processes. Canopy nitrogen stress can be measured by the red-edge band located at 0.715 μm. Remote sensing of NPV requires three SWIR bands between 2.025 – 2.230 μm that allow for calculation of the Cellulose Absorption Index. Split TIR bands are needed to measure land surface temperatures, which allow for estimation of actual ET and CO₂ assimilation by crop canopies. Future agricultural and land cover monitoring systems should have 12-bit or better quantization, signal-to-noise ratios of > 500, pixel sizes < 60 m, and temporal resolutions of < 7 days to ensure cloud-free scenes and to capture critical crop growth stages.

Presentation: 2011_Poster_Serbin_244_281.pptx (2614k)

Presentation Type:  Poster

Session:  Science in Support of Decision Making   (Wed 10:00 AM)

Associated Project(s): 

• Related Activity

Poster Location ID: 244